© 2015. Published by The Company of Biologists Ltd | Biology Open (2015) 4, 1270-1280 doi:10.1242/bio.013540
RESEARCH ARTICLE
A zebrafish model of inflammatory lymphangiogenesis
ABSTRACT Inflammatory bowel disease (IBD) is a disabling chronic inflammatory disease of the gastrointestinal tract. IBD patients have increased intestinal lymphatic vessel density and recent studies have shown that this may contribute to the resolution of IBD. However, the molecular mechanisms involved in IBD-associated lymphangiogenesis are still unclear. In this study, we established a novel inflammatory lymphangiogenesis model in zebrafish larvae involving colitogenic challenge stimulated by exposure to 2,4,6-trinitrobenzenesulfonic acid (TNBS) or dextran sodium sulphate (DSS). Treatment with either TNBS or DSS resulted in vascular endothelial growth factor receptor (Vegfr)dependent lymphangiogenesis in the zebrafish intestine. Reduction of intestinal inflammation by the administration of the IBD therapeutic, 5-aminosalicylic acid, reduced intestinal lymphatic expansion. Zebrafish macrophages express vascular growth factors vegfaa, vegfc and vegfd and chemical ablation of these cells inhibits intestinal lymphatic expansion, suggesting that the recruitment of macrophages to the intestine upon colitogenic challenge is required for intestinal inflammatory lymphangiogenesis. Importantly, this study highlights the potential of zebrafish as an inflammatory lymphangiogenesis model that can be used to investigate the role and mechanism of lymphangiogenesis in inflammatory diseases such as IBD. KEY WORDS: Zebrafish, Inflammation, Lymphatic, Inflammatory bowel disease
INTRODUCTION
Crohn’s disease and ulcerative colitis are characterised as chronic inflammatory disorders of the gastrointestinal tract and together are known as inflammatory bowel disease (IBD). It is well accepted that the dysregulation of normally controlled immune responses to gut microbiota, as well as genetic susceptibility and other environmental factors contribute to the pathogenesis of IBD (Zhang and Li, 2014), however the aetiology of IBD is still not well understood. Increased intestinal lymphatic vessel density is observed in IBD patients (Fogt et al., 2004; Geleff et al., 2003; Kaiserling et al., 2003; Pedica et al., 2008; Rahier et al., 2011) and in mouse models of IBD (D’Alessio et al., 2014; Ganta et al., 2010; Jurisic et al., 2013). Although increased lymphangiogenesis is proposed to occur in response to the decreased mesenteric lymphatic function observed in 1
Department of Molecular Medicine & Pathology, School of Medical Sciences, 2 University of Auckland, Auckland 1142, New Zealand. Department of Molecular Genetics and Microbiology, Duke University Medical Center, Durham 27710, USA. 3 Australian Regenerative Medicine Institute, Monash University, Clayton, Victoria 3800, Australia. *Author for correspondence (
[email protected]) This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution and reproduction in any medium provided that the original work is properly attributed.
Received 19 July 2015; Accepted 5 August 2015
1270
IBD patients (Heatley et al., 1980; Van Kruiningen and Colombel, 2008; Van Kruiningen et al., 2014), the mechanism by which these lymphatic vessels form is still unclear. The growth of lymphatic vessels during early lymphatic development largely involves the sprouting of lymphatic endothelial cells from the veins (Srinivasan et al., 2007; Yaniv et al., 2006) and requires vascular endothelial growth factor 3 (VEGFR-3) signalling (Irrthum et al., 2000; Karkkainen et al., 2000, 2001; Veikkola et al., 2001). The binding of extracellular VEGF ligands, in particular VEGF-C and VEGF-D, activates VEGFR-3 (Karkkainen et al., 2004; Kukk et al., 1996). Reducing inflammatory lymphangiogenesis using a VEGFR-3 blocking antibody, or decreasing lymphatic function through the down regulation of FoxC2, a gene required for lymphatic homeostasis, significantly increases inflammatory oedema and inhibits disease resolution in mouse models of IBD, suggesting a protective role of intestinal lymphatics in IBD (Becker et al., 2015; Jurisic et al., 2013). In agreement, a recent study has shown that enhancing lymphangiogenesis and lymphatic function, by adenoviral induction of VEGF-C, alleviates experimental chronic intestinal inflammation by increasing inflammatory cell mobilisation and bacterial antigen clearance (D’Alessio et al., 2014). Although these studies suggest there is potential in treating IBD by improving intestinal lymphatic function, further studies are required to determine the mechanism of IBD-associated lymphangiogenesis. While the VEGF-C/VEGFR-3 signalling pathway has been associated with mammalian IBD-associated lymphangiogenesis (D’Alessio et al., 2014; Jurisic et al., 2013), the significance of other VEGFR signalling pathways implicated in inflammatory lymphangiogenesis, such as VEGF-A/VEGFR-2 (Cursiefen et al., 2004; Halin et al., 2007; Kataru et al., 2009; Kunstfeld et al., 2004; Tan et al., 2013) and VEGF-D/VEGFR-3 (Baluk et al., 2005; Huggenberger et al., 2010; Kataru et al., 2009; Tan et al., 2013) is still undetermined. In addition, the cells and tissues providing the prolymphatic VEGFs that mediate IBD-associated lymphangiogenesis remain to be identified. In IBD patients, increased levels of lymphatic growth factors VEGF-A, VEGF-C, and VEGF-D have been reported in serum and mucosa culture supernatants (Algaba et al., 2013; Bousvaros et al., 1999; D’Alessio et al., 2014; Duenas Pousa et al., 2007; Kanazawa et al., 2001; Pousa et al., 2008). Peripheral blood mononuclear cells (Griga et al., 1999a,c), intestinal mucosa (Griga et al., 1999a, 2002), adipose tissue (Schaffler et al., 2006), and fibroblasts (Beddy et al., 2004) have all been suggested to provide VEGF-A in IBD and may be particularly important for IBDassociated angiogenesis. However, the cellular sources of VEGF-C (and VEGF-D) required for IBD-associated lymphangiogenesis are still not clear. In other inflammatory situations such as chronic airway inflammation, skin inflammation, keratitis, and peritonitis, macrophages express pro-lymphatic growth factors, such as VEGFA, VEGF-C, and VEGF-D, and macrophage depletion results in reduced inflammatory lymphangiogenesis (Baluk et al., 2005; Cursiefen et al., 2004; Harvey and Gordon, 2012; Kataru et al.,
Biology Open
Kazuhide S. Okuda1, June Pauline Misa1, Stefan H. Oehlers2, Christopher J. Hall1, Felix Ellett3, Sultan Alasmari3, Graham J. Lieschke3, Kathryn E. Crosier1, Philip S. Crosier1 and Jonathan W. Astin1,*
2009; Kim et al., 2009; Kubota et al., 2009; Maruyama et al., 2005; Tan et al., 2014; Zhang et al., 2007). A recent study has shown that neutrophils can stimulate skin inflammation-associated inflammatory lymphangiogenesis by increasing VEGF-A bioavailability, and also by secreting VEGF-D (Tan et al., 2013). Intestinal epithelial cells have also been shown to express VEGF-C (Joory et al., 2006). Therefore, it is possible that macrophages, neutrophils, and intestinal epithelial cells may contribute to IBD-associated lymphangiogenesis by secreting VEGF-A, VEGF-C, and VEGF-D. The availability of zebrafish transgenic lines that mark lymphatic vessels (Gordon et al., 2013; Okuda et al., 2012; van Impel et al., 2014), leukocytes (Ellett et al., 2011; Hall et al., 2007), and the intestinal epithelial cells (Her et al., 2004) makes the zebrafish model an ideal platform to uncover the in vivo interaction between these cell populations and lymphatic vessels in IBD. Zebrafish larvae treated with the colitogenic agents 2,4,6-trinitrobenzenesulfonic acid (TNBS) or dextran sodium sulphate (DSS) develop intestinal inflammation with IBD-like characteristics including: (1) the requirement for microbiota to trigger inflammation; (2) responsiveness to anti-inflammatory medications; (3) increased expression of pro-inflammatory cytokines; (4) increased leukocyte recruitment to the intestine; (5) enlarged intestinal lumen and smoothening of microvilli (Fleming et al., 2010; Oehlers et al., 2012, 2011; Yang et al., 2014). However, as the zebrafish are immersed in the colitogenic agents, inflammation is not restricted to the intestine in this IBD-model. Zebrafish are also gaining standing as a model for lymphatic development and, importantly, the requirement for Vegfr3/Vegfc signalling in lymphatic development is conserved in zebrafish (Hogan et al., 2009; Kuchler et al., 2006; Le Guen et al., 2014; Okuda et al., 2012; Yaniv et al., 2006). Combining existing zebrafish IBD and lymphatic models may therefore provide a novel platform for IBD-associated lymphangiogenesis research. In this study, we show that intestinal inflammation triggered by exposure of zebrafish larvae to TNBS or DSS resulted in Vegfr-dependent lymphangiogenesis, specifically in the zebrafish intestine. Furthermore, we show that zebrafish leukocytes express vascular growth factors and that macrophages are required for inflammatory lymphangiogenesis following colitogenic challenge. RESULTS Colitogenic challenge using TNBS or DSS stimulates development of lymphatic sprouts in the intestine
The aim of this study was to establish an inflammatory lymphangiogenesis model in zebrafish. To do this, we utilised the lyve1:DsRed2;kdrl:EGFP compound transgenic line which labels both the lymphatic/venous (lyve1) and blood (kdrl) vasculature and can therefore be used to differentiate the lymphatic vessels from the blood vessels (Okuda et al., 2012). Previous characterisation of this line has revealed that zebrafish develop an intestinal lymphatic (IL) network (summarised in supplementary material Fig. S1) (Okuda et al., 2012). Intestinal lymphatic sprouts (ILSs) (previously described as lymphatic branches) grow between the major intestinal lymphatic vessels and are rare on the left side of the zebrafish intestine at 7 days post-fertilisation (dpf). From 7 dpf, the number and length of the ILS steadily increases until 15 dpf, where they are hypothesised to contribute to the web-like IL that forms across the intestinal bulb (Okuda et al., 2012). The lack of ILSs on the left side of the larvae at 7 dpf provides a physiologically-relevant opportunity to identify and quantify intestinal lymphatic network expansion in intestinal disease. When intestinal inflammation was induced using either TNBS or DSS in 3 dpf lyve1:DsRed2;kdrl:EGFP embryos, the number and
Biology Open (2015) 4, 1270-1280 doi:10.1242/bio.013540
total length of the ILSs were increased at 7 dpf when compared with untreated larvae (Fig. 1). Increased lymphangiogenic activity was intestine-specific, as no ectopic lymphatic vessels were observed in the trunk (Fig. 1A‴-C‴) or the head (data not shown). The lymphatic vessels induced following TNBS or DSS treatment predominately grew over the outer surface of the intestinal epithelium. However, there were rare examples where these ILS grew towards the swim bladder, dorsal to the intestine. Blood vessel development in the intestine also appeared normal in TNBS and DSS-treated larvae (Fig. 1A-C). To show that the increased ILS formation was inflammation-dependent, we suppressed inflammation in TNBS or DSS treated larvae by co-administration of the anti-inflammatory drug 5-aminosalicylic acid (5-ASA) which is used to treat IBD (Rutgeerts et al., 2009). With 5-ASA co-treatment, the number and total length of the ILSs was reduced when compared with larvae exposed to colitogenic agents alone (Fig. 1D-G), showing that TNBS/ DSS-driven lymphangiogenesis in the intestine is associated with inflammation. We therefore termed this model zebrafish intestinal inflammatory lymphangiogenesis (IIL). 5-ASA reduces neutrophil and macrophage recruitment to the intestine
Given that macrophages and neutrophils are known to contribute to inflammatory lymphangiogenesis in mammalian models (Tan et al., 2014), we next investigated whether reduction of zebrafish IIL following 5-ASA treatment was associated with decreased immune cell recruitment to the intestine. Using the neutrophil-specific lysozyme C promoter driven Tg(lyz:EGFP) transgenic line (Brannon et al., 2009; Hall et al., 2007) we confirmed our previous study showing that TNBS exposure results in an increase in neutrophil recruitment to the zebrafish intestine and that treatment with 5-ASA suppresses this (Oehlers et al., 2011) (see supplementary material Fig. S2). To establish whether TNBS treatment also altered macrophage recruitment to the intestine, larvae expressing the macrophage-specific macrophage expressed gene 1 promoter driven Tg(mpeg1:EGFP) transgene (Ellett et al., 2011) were treated at 3 dpf with TNBS and the number of mpeg1-positive, macrophagelineage cells in the intestine assessed at 7 dpf. The number of macrophages in the intestine of TNBS-treated larvae increased 2.2-fold when compared with untreated larvae (untreated, 34±12; TNBS-treated, 73±18) (Fig. 2A,B,H). Confocal live imaging of double transgenic mpeg1:EGFP;intestinal fatty acid binding protein (i-fabp):RFP (Her et al., 2004) larvae demonstrated that macrophages accumulated around the outer surface of the i-fabpexpressing intestinal epithelial cells following TNBS treatment (Fig. 2D-G). The total number of EGFP-positive macrophages in mpeg1:EGFP larvae, quantified by fluorescence activated cell sorting (FACS), increased 1.5-fold following TNBS treatment (untreated, 68/100,000±10; TNBS-treated, 99/100,000±10) (Fig. 2I). From this data we conclude that TNBS induces a systemic inflammatory response that is particularly severe in the intestine. Finally, 5-ASA suppressed the recruitment of macrophages to the zebrafish intestine in larvae exposed to TNBS; larvae treated with 5-ASA+TNBS had less intestinal macrophages when compared with TNBS-treated larvae (Fig. 2A-C,H). Taken together these results show that 5-ASA co-treatment reduces TNBS-mediated inflammation in the zebrafish intestine. Zebrafish macrophages and neutrophils express lymphatic growth factors
RT-PCR was used to investigate whether zebrafish larval macrophages express lymphatic growth factors. Macrophage1271
Biology Open
RESEARCH ARTICLE
Biology Open (2015) 4, 1270-1280 doi:10.1242/bio.013540
Fig. 1. Colitogenic challenge is associated with increased intestinal lymphangiogenesis. (A-C) Lateral images of lyve1:DsRed2;kdrl:EGFP larvae at 7 dpf (A), treated with TNBS (B) or DSS (C). Asterisks indicate intestinal lymphatic sprouts (ILSs). A′-C′ shows the left intestinal vasculature and A‴-C‴ shows the trunk vasculature in the DsRed channel only. A″-C″ are schematic diagrams of arteries (red), veins (blue), lymphatic vessels (green) and ILS (magenta) of A-C. (D-G) Quantification of ILS number (D,E) and total ILS length (F,G) in TNBS (D,F) and DSS (E,G) treated larvae compared with untreated, 5-aminosalicylic acid (5-ASA), and TNBS/DSS+5-ASA (n≥20). Error bars, ±s.d. *P
